Comprehensive data protection for all workloads
Post Reply
perjonsson1960
Veteran
Posts: 463
Liked: 47 times
Joined: Jun 06, 2018 5:41 am
Full Name: Per Jonsson
Location: Sweden
Contact:

requires mount server with data deduplication feature enabled

Post by perjonsson1960 »

Perhaps I spoke too soon when I said that it works perfectly in v11a... I did an active full backup of our file server failover cluster from Friday to Saturday (approx. 18 TB). Everything looked fine, it was "Success" all over the board. However, when I do a restore of a certain folder from that active full, I get about half the files restored. It says "Files to restore: 14, folders to restore: 1". And when it is finished it says "Files restored: 8, folders restored: 1". No errors reported.

Whenever I am about to do a restore from the cluster I get this message: "Copying and restoring files from deduplicted volumes requires that the mount server associated with the source backup repository uses Windows Server 2012 R2 or later with data deduplication feature enabled."

What does this mean? I know that the file server cluster has the dedupe feature enabled in order to save space, but as far as I know, that is the only server that uses the dedupe feature. The VBR repositories do not. We are using VBR's own inline dedupe and fast clone in the backup jobs, and I have been told NOT to use both MS dedupe feature and VBR's at the same time.

#SPLIT from ReFS 2019 vs. 2022
Mildur
Product Manager
Posts: 8735
Liked: 2296 times
Joined: May 13, 2017 4:51 pm
Full Name: Fabian K.
Location: Switzerland
Contact:

Re: requires mount server with data deduplication feature enabled

Post by Mildur »

Hi Per Jonsson

I split this question from the other topic. It's not related to the backup repository target topic between upgrades.

When you want to restore guest files from a windows machine which had volumes with windows data deduplication enabled, a mount server with windows data deduplication feature is required.

You start the Guest OS file restore. The backup server uses the mount server assigned in the backup repository properties to mount the backup of this machine in C:\VeeamFLR. The restore Explorer connects to that folder to read and restore the content. If the mount server doesn't have the windows data deduplication feature enabled, the mount server cannot read those backed up volumes with windows data deduplication enabled.

In Short:
- Mount Server in your environment must have the same or higher version than the protected VMs.
- Mount Server in your environment must have windows data deduplication feature enabled.

Best,
Fabian
Product Management Analyst @ Veeam Software
perjonsson1960
Veteran
Posts: 463
Liked: 47 times
Joined: Jun 06, 2018 5:41 am
Full Name: Per Jonsson
Location: Sweden
Contact:

Re: requires mount server with data deduplication feature enabled

Post by perjonsson1960 »

Do I only have to install the dedupe feature in the repo servers, or do I also have to enable the dedupe feature on the backup volumes? As I said, I have understood that you should not use the MS dedupe feature together with VBR's dedupe, and we are using VBR's dedupe in every job.

Is the backup I did of the cluster useless? And what about all the other backups that reside on the same volumes? We have all our 300+ VMs and physical machines backed up on the same SOBR as where the file cluster is backed up.
Mildur
Product Manager
Posts: 8735
Liked: 2296 times
Joined: May 13, 2017 4:51 pm
Full Name: Fabian K.
Location: Switzerland
Contact:

Re: requires mount server with data deduplication feature enabled

Post by Mildur »

Only on the mount server, that's what the error message is telling you.
Copying and restoring files from deduplicated volumes requires that the mount server associated with the source backup repository uses Windows Server 2012 R2 or later with data deduplication feature enabled.
You can check the assigned mount server in your repository settings:
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

This requirement is also documented in our user guide:
https://helpcenter.veeam.com/docs/backu ... uplication

Best,
Fabian
Product Management Analyst @ Veeam Software
perjonsson1960
Veteran
Posts: 463
Liked: 47 times
Joined: Jun 06, 2018 5:41 am
Full Name: Per Jonsson
Location: Sweden
Contact:

Re: requires mount server with data deduplication feature enabled

Post by perjonsson1960 »

The mount server is the same as the repo server, which is the default setting, I think.

We have three different repo servers, an old server where there is a sealed extent (2019), an old server where the backup copies are written (2019), and a new server where all the primary backups are written (2022). None of them has the MS dedupe feature installed.

But you didn't answer my questions. I am very worried about this, so I would appreciate answers.

1. Do I only need to install the MS dedupe feature, or do I also have to enable the MS dedupe feature on all the VBR backup volumes associated with the mount server?
2. Is the current full backup of the file cluster useless, and needs to be redone after the dedupe feature is installed (or even enabled) in the mount server, or will it be usable after the dedupe feature is installed (or even enabled)?
3. What happens with all the other backups that are already written to the repos if the dedupe feature is installed (or even enabled) on the VBR backup volumes?
4. Is it really recommended to use both MS dedupe and VBR's dedupe on the same volume, if the MS dedupe is enabled on the VBR backup volumes?

In short, what do I need to do in order to be able to actually restore from the backups that we do?
Mildur
Product Manager
Posts: 8735
Liked: 2296 times
Joined: May 13, 2017 4:51 pm
Full Name: Fabian K.
Location: Switzerland
Contact:

Re: requires mount server with data deduplication feature enabled

Post by Mildur »

1. Do I only need to install the MS dedupe feature, or do I also have to enable the MS dedupe feature on all the VBR backup volumes associated with the mount server?
Only enable the feature. Do not enable windows data deduplication for the volumes on the backup repository server.
2. Is the current full backup of the file cluster useless, and needs to be redone after the dedupe feature is installed (or even enabled) in the mount server, or will it be usable after the dedupe feature is installed (or even enabled)?
No. It's not useless. The VM (or physical machine?) was backed up entirely. You only need data deduplication feature on your mount server to read the deduped blocks in the backup file. It is not used to backup those volumes.
3. What happens with all the other backups that are already written to the repos if the dedupe feature is installed (or even enabled) on the VBR backup volumes?
Nothing. After the feature is enabled on the server, nothing happens to the server volumes. And you don't have todo anything. Just enable the data deduplication feature. Don't enable deduplication on the volumes.
4. Is it really recommended to use both MS dedupe and VBR's dedupe on the same volume, if the MS dedupe is enabled on the VBR backup volumes?
We never recommend to use windows data deduplication for a backup repository volume. Use reFS instead.
But this is a question for a new topic. Not relevant for the issue in this topic.

Best,
Fabian
Product Management Analyst @ Veeam Software
perjonsson1960
Veteran
Posts: 463
Liked: 47 times
Joined: Jun 06, 2018 5:41 am
Full Name: Per Jonsson
Location: Sweden
Contact:

Re: requires mount server with data deduplication feature enabled

Post by perjonsson1960 »

Thank you very much!
I feel a bit calmer now! ;-)
So, I only need to enable the dedupe feature in the mount servers, and not touch the backup volumes. Good!

Kind regards,
PJ
Post Reply

Who is online

Users browsing this forum: Semrush [Bot] and 117 guests